Component Supplier

Member Since 2016

LEARN MORE

Radiant Vision Systems provides the leading test & measurement systems and software for characterization and automated visual inspection of displays, backlit components, light sources, and device surfaces. Our imaging photometers and colorimeters test millions of devices in production settings worldwide.

Content Filed Under:

Industry:
N/A

Application:
N/A

CP Industries Delivers Turnkey Test Systems for Head-Up Display Windshield Glass Using ProMetric® Imaging Photometers and TT-HUD™ Software

POSTED 04/20/2021

 | By: Radiant Vision Systems, Case Study

About CP Industries

CP Industries is a full-service tooling company headquartered in Granger, Indiana, U.S.A., with facilities throughout the United States and Mexico. The company is a fifth-generation, family-owned business that serves manufacturers with end-to-end assembly and test systems for glass, trim, thermoform, and injection-molded components. The company’s fixtures are custom-built to meet exact customer requirements, with all product design, assembly, machining, installation, and training provided by CP Industries to minimize time, cost, and complexity for their customers.

Since the company’s foundation in 1937, CP Industries has built a decades-long history delivering solutions to the automotive industry. Applying expertise in tooling, gaging, and fixturing for automotive glass, the company has worked extensively with glass manufacturers as well as major automotive OEMs and tiered suppliers to ensure quality glass components for vehicles. With the advent and growth of head-up displays (HUD) in the automotive market, CP Industries was well-positioned to address the need for windshield testing as an integral component of HUD systems. However, reporting an objective measure of HUD visual quality according to automotive industry and OEM specifications is challenging for traditional inspection equipment and software. As the company began developing its new HUD test system, CP Industries looked to optimize components that would address the advanced requirements of HUD glass inspection and solve more customer challenges out of the box.


Requirements for HUD Glass Testing

Automotive HUD systems project virtual images onto the vehicle’s windshield, which are visualized through the glass at a perceived distance of several meters. This enables drivers to view information like speed and navigation within their line of sight while driving, reducing the need to take their eyes off the road. Because the windshield acts as an optical component of the HUD system, the visual performance of HUDs can be dramatically impacted by the glass through which information is projected and viewed. Variations in windshield glass form, lamination, layer thickness, and wedge angle may result in visible defects that appear as dimensional distortion or ghosting (duplication) of HUD virtual images when viewed through the glass. Requirements for HUD testing are based on industry-standard test parameters and formulas in addition to OEM-specified requirements that must be met by suppliers throughout the HUD ecosystem, from glass makers to projection system manufacturers.

Example of testing visual HUD performance through the windshield, using a camera system for automated visual inspection.

“Our glass customer receives test requirements from the automotive OEM, and those requirements are changing all the time,” states Cruz Palafox, Project Engineer at CP Industries. “We have also worked in cases where the customer doesn’t have any specifications up front, and we have to try to develop something to meet what they’re asking for.”

Adding to the complexity of HUD testing is the number of visual criteria that must be measured for a complete inspection. HUD glass is tested for several primary defect categories—for example, Distortion and Ghosting. Test requirements for each defect category may vary, but fundamentally include a test image (usually a line or dot pattern) that is projected through the glass, and formulas that use coordinate points in the image to calculate geometric measurements of deviations in the projected pattern from the ideal pattern. Each defect category has several qualities that must be evaluated, and each quality is measured using a specific formula and method (specified in industry or OEM requirements). For example, a complete Distortion evaluation for HUD glass inspection may require 10-20 calculations to evaluate all visual qualities to account for the range of potential deviations within the projected test image.

In the end, a complete HUD glass inspection will output a measurement (data point) for each quality evaluated. Thresholds can be used to pass and fail the glass during automated testing based on the type or severity of defects detected, as indicated by the measured values.  

Defect Category

Distortion

Ghosting

Test Image

(example)

Defect (example)

A black background with white dotsDescription automatically generated with medium confidence

Horizontal trapezoid distortion

Single ghost, with separation

Resulting HUD Image

(example)

Qualities evaluated* (example)

  • Size (horiz./vert.)
  • Aspect ratio
  • Rotation
  • Translation (horiz./vert.)
  • Angular displacement (horiz./vert.)
  • Trapezoid (horiz./vert.)
  • Smile (horiz./vert.)
  • Magnification (horiz./vert.; avg., max.)
  • Number of ghosts
  • Distance to primary ghost
  • Average ghost (horiz./vert.)
  • Max. ghost (horiz./vert.)

* Performed for each eyebox position.

Optimizing Efficiency with Imaging and Software

Inspecting HUD glass for dimensional defects requires both a spatial understanding of the HUD projection (an image captured by a camera) and the ability to locate and compare several coordinate points across a projected test image (image analysis typically performed by software).

“We started using machine vision cameras, which is how we built a foundation of working with cameras,” explains CP Industries Project Engineering Manager, Connor Cassady. “Customers wanted us to be able to view different things on glass, from barcode labels to QR codes to components that are mounted on glass.” As the company moved toward the development of HUD test systems, machine vision cameras continued to provide imaging benefits for visualizing a complete HUD projection through the windshield. This image could then be analyzed in machine vision software to provide necessary information like coordinate locations of test image features (lines and dots) to be used in calculations required for Distortion and Ghosting inspection.

“We were capturing dimensional data from the HUD test patterns, and the captured data was sent back to Excel® into a report,” states Andrew Early, Control Engineer at CP Industries. But unlike reading a barcode or detecting a component, processing images and applying analyses for the breadth of criteria in a single HUD glass inspection proved taxing for the machine vision system. As Early explains, “The biggest issue we experienced with the initial machine vision cameras was limited memory space and the amount of data that could be output at once. This meant our programs had to be limited to a certain number of tools.”

Throttling and readdressing data output from the machine vision system was a potential solution but added to development work. “I had to basically break out my data output, which took more time,” says Early. Data management became an even bigger issue if the customer required images of each HUD analysis to be saved and stored for verification and reporting purposes.

Once data was output, a final challenge needed to be addressed—applying the unique formulas to the data to calculate the measurement values for HUD glass evaluation. As Early recalls, “Originally, we did that work post-process, in the Excel side. We developed the inspection test program ourselves in an Excel format, including all formulas for HUD glass inspection. But that increased our cycle time.” Further, if an operator at the customer site were to unlock the Excel file after test system delivery, unintended edits or resorting of the data or formulas could cause errors in the output measurement values of the test report.

As the engineering team at CP Industries began work on their seventh-generation HUD glass test system—the HUDSON test system—optimizing efficiency was key. By replacing the original machine vision imaging system and software, the team could streamline data management, reduce development time and complexity, and ultimately pass along shorter lead times, cycle times, and usability improvements to their customers.

The HUDSON test system from CP Industries addresses complete customer requirements for HUD glass inspection and can be incorporated as part of a fully automated production line or used in offline measurement.

CP Industries selected a scientific HUD measurement system from Radiant Vision Systems to replace machine vision cameras and software in its new HUDSON system. A provider of test and measurement for automotive light sources and displays, Radiant has built a fundamental understanding of HUD measurement criteria through the application of photometric and colorimetric imaging systems for HUD testing across OEM and supplier tiers. Radiant’s HUD measurement solutions incorporate standard test criteria specified by the Society of Automotive Engineers (SAE J1757-2) as well as features developed around unique customer application requirements. Radiant’s systems are applied to test photometric qualities like luminance, chromaticity, and contrast, as well as dimensional checks for distortion, warping, and ghosting. High-resolution, scientific-grade ProMetric® imaging systems designed for these analyses are equipped to handle significant data processing and output demands, and provide on-board image processing and advanced calculations for image-based measurement.

CP Industries integrated Radiant’s ProMetric Y Imaging Photometer into its HUDSON test system to handle the image capture and processing for HUD glass inspection. Additionally, data from the ProMetric system is analyzed using Radiant’s TT-HUD™ Software platform, a test suite and user interface developed specifically for automated visual inspection of automotive HUDs. TT-HUD provides all data capture and analysis functions, including built-in calculations to measure the qualities of HUD projections based on standard formula.

The ProMetric® Y Imaging Photometer captures images of complete HUD projections, and provides on-board image processing and analysis using a suite of HUD measurement tests in TT-HUD™ Software.

“I was having to program all of the tests from scratch that Radiant was providing out of the box. Whenever the customer needed to make a small change, it drastically impacted the programming I had done, and required me to restructure everything,” recalls Early. “The Radiant system does all image and data processing internally, rather than having to manage the calculations of the data in post-processing. The flexibility and the adaptability of Radiant to different testing requirements was another plus.”

With imaging processing and calculations performed by the Radiant system, the data sent to the customer’s test report is the final output data, with all formulas pre-applied. Not only does this reduce the risk of errors if the Excel report is adjusted, but there are no additional processes that need to be developed to manage data, reducing both cycle times per test and development time for the overall system.

“One of our main goals was cycle time. Since adding the Radiant system, we definitely see an improvement in speed,” says Palafox. “This gives us a better lead time on our end to provide a machine to a customer—from around 24 weeks for a system to around 14.”

Looking ahead, Radiant’s expertise and range of test functions can be applied to expand and develop test systems to address a range of new criteria in HUD glass testing. This enables CP Industries to meet a more diverse set of needs and take on new projects for its automotive and glass customers in the future.

Palafox explains, “We have the ability to do certain measurements that we haven’t done before, but with Radiant, we can easily do. Before, we could perform basic tests for ghosting and image rotation—but not dynamic geometric measurement. The Radiant camera seems like it’s capable of doing a lot more if we get to the point of having to measure even more specifications. We can handle that now, whereas we were much more limited with the other camera options.”

Optimizing Repeatability with Robotics and Optics

The visual performance of a HUD projection can be significantly impacted by the angle and position from which it is viewed. Eyebox is an important factor of HUD glass testing, accounting for the volume of space where the driver’s eyes could be located to visualize the HUD projection. Imaging system position and rotation may be specified in HUD test requirements to account for eyebox positions, to simulate the possible heights and positions of a driver. These variations could influence the appearance of defects like Distortion and Ghosting caused when light from the HUD projection is reflected and refracted through the various windshield layers and influenced by the wedge angle (the angle between the inner and outer layers of the windshield glass).

Illustration of the HUD eyebox—within this area, the HUD virtual image is completely visible through the windshield.

For CP Industries, meeting test requirements for angle and position meant that they would need to build motion into the HUD glass test system.

“We started with three stationary-positioned machine vision cameras and a HUD projector, which moved during testing,” states Palafox. However, moving the projector proved problematic for repeat inspections.

“A HUD projector that’s integrated in the car is intended to be adjusted for maybe 10 seconds every month,” explains Early. “Whereas in a production setting, it would be moving up and down every 10 seconds, with its motor running 24 hours a day.” This movement quickly wears out the HUD projection system.

In the design of their HUDSON test system, the CP Industries team resolved this issue by eliminating the HUD projector entirely. Instead, they added an LED to simulate projection when cast through a custom metal grating, and robotics that would control the position of the inspection camera.

“We create a plate with a test pattern (that the customer usually specifies), and behind that are LEDs that shine through the plate to project the pattern through the glass,” explains Palafox. “The LEDs that we put on our HUDSON machine have a lifespan of up to 10 years. And now the camera is moving instead of the projector. That’s a big benefit for the glass companies compared to having to buy a new projector every one-to-two months.”

Using the Radiant ProMetric Imaging Photometer, the team was also able to reduce equipment in the HUDSON system, eliminating components previously needed to ensure test system performance. As Cassady explains, “We used to have a luminance meter in the test system to make sure the LEDs were emitting enough light for the test projection. But now we don’t need that.” As a photometric camera, the ProMetric Y captures absolute luminance values (cd/m2) in addition to dimensional HUD inspection data.

        

During HUD glass inspection, LED light is cast through the HUDSON’s projector plate (left) and through the windshield under test, emulating a HUD projector. The target plate (right) is used similarly to produce an ideal HUD projection for system calibration.

With the movement and rotation of so many components, the HUDSON test system needed to be designed to account for variables that could impact misalignment of the camera system, optics, and the robotics in relation to the HUD glass and projection. The CP Industries team added an automated calibration routine to the HUDSON system using a target plate above the camera, illuminated by an LED to create a golden sample of the ideal HUD projection test image.

“Calibration ensures the robot isn’t worn, the camera isn’t damaged, the machine hasn’t shifted or changed,” Early explains. “The target plate above is used to calibrate at the beginning of each test run. If there’s any doubt that the machine isn’t providing accurate data, the customer can go back to that calibration plate and run a calibration verification.”

Adjusting to positional changes in both testing and calibration meant that the imaging system would need to adapt quickly to new optical parameters, particularly camera focus. HUD virtual images are projected several meters out from the driver position to appear in-line with the environment outside the vehicle. Taking additional eyebox testing requirements into account, the focus settings of the imaging system would have to change for different HUD projection depths as robotics move the camera to different locations. Regardless of these changes, the imaging system must ensure test images are captured clearly and accurately for consistent measurement performance at all eyebox locations.

Integrating the Radiant ProMetric Y Imaging Photometer into their HUDSON system, CP Industries was able to realize the benefits of the camera’s electronic lens for accurate and repeatable focus-setting. During automated measurement, the ProMetric Imaging Photometer’s focus and aperture settings can be adjusted remotely via software and adapted dynamically to different test positions. The electronic lens also enables quick, hands-free calibration of the test system using the HUDSON’s calibration target plate.

“With the electronic lens, there’s less concern when you’re moving the camera around,” states Early. “There are of course vibrations from the robot, maintenance technicians might clean the system, etc. But you don’t have to worry about someone moving that lens and shifting your focal point, which drastically affects your measurements.”

 

Electronic lenses enable quick, remote adjustment of the camera’s focus and aperture via software to match exact HUD projection distances. ProMetric Imaging Systems are calibrated for each lens across a range of working distances and aperture settings.

“Using the machine vision cameras, we had a bunch of trouble with focus and finding the correct distance to the projection. Then, we had to go in and adjust lenses manually,” adds Palafox. “If you moved the lens even a little bit, you could lose focus, and that would be a big hassle. Radiant’s electronic lens saves us time during setup of the test system as well—it’s a big advantage in bringing down lead time.”

Combining electronic lenses with fully automated test sequencing in TT-HUD, Radiant ProMetric imaging systems provide an ideal in-line HUD measurement solution for pass/fail analysis across multiple visual inspection parameters. With built-in API and SDK, TT-HUD also supports integration with robotics and automation equipment or manufacturing systems, providing ultimate flexibility in test systems development to meet diverse customer needs in the future.

Summary

Dedicated to innovation, CP Industries continues to optimize its HUD glass test systems to provide the most comprehensive solutions for its customers. From meeting complete HUD test specifications, to developing custom test routines, to designing automation to fit each unique production line, CP Industries’ systems are engineered to do more so customers can focus on producing the best glass products for HUD applications.

With the help of HUD measurement technology from Radiant Vision Systems, CP Industries’ turnkey solutions now leverage powerful functions to address the gamut of HUD inspection tasks, including HUD-focused software tools, multi-functional photometric imaging, electronic optics, robotics, and more. By replacing machine vision systems with Radiant’s ProMetric Imaging Photometer and TT-HUD Software in its latest HUDSON test system, CP Industries has streamlined its development, reduced equipment, and improved functionality. These optimizations have allowed the company to pass along improved lead time, testing speed, cost-efficiency, and capability to its customer base, illustrating why CP Industries is chosen by so many for tooling systems that support automotive manufacturing and beyond.

“Radiant and CP Industries build on each other’s knowledge to develop HUD testing. As opposed to working with the previous machine vision systems—which relied on us for all the fundamental HUD test knowledge—Radiant has that background to help us understand what else we can do in HUD testing. It’s not all on us.”

  • Cruz Palafox, Project Engineer, CP Industries

Learn more about CP Industries: www.cpind.com